Social media:
An effective instrument for
an antisocial movement
Amsterdam University of Amsterdam
Name of Supervisor Name of Second Reader
Mapping the rise of the German
anti-Islam movement Pegida on
24th June 2016
MA New Media and Digital Culture
dhr. prof. dr. R.A. Rogers mw. drs. L.C. van der Velden
Abstract
Digital media are often studied for the way they change the formation of protest movements since they are said to function as organizing agents and make hierarchical structures unnecessary. In combination with a strong personalization aspect, those new types of protest movements are described as connective action networks that have been so far only researched in relation to leftist or slacktivism protests. Therefore I used the German right-wing and anti-Islam movement Pegida as an example in order to evaluate to what extent a right-wing logic of connective action can be determined. To analyse the dynamics of Pegida on Facebook, a digital methods approach was chosen including a content analysis and a page-like network analysis of the main PEGIDA Facebook page. The findings suggest that memes trigger the most engagement on the Facebook page and enable a strong personalization of the content since they are easy-to-adapt personally. On the other hand Pegida also shows some hierarchical structures and organizational coordination what aligns more with the notion of traditional collective action movements. Therefore Pegida can be categorized as a hybrid version of a traditional collective action and a self-organizing connective action network or, in other words, as organizationally enabled network where loose organization is involved but the personalization of the content plays a primary role.
Keywords
Social Network Analysis, Facebook, Netvizz, Pegida, anti-Islam movements, echo chamber linking, right-wing populism, right-wing logic of connective action
Table of Content
I.
Introduction
... 1
II.
Literature Review
... 3
2.1 Notions of Extremist Movements Online ... 3
2.1.1 Echo Chamber Linking and the Creation of a Network of Hate ... 4
2.1.2 Recruitment, Mobilisation and Radicalisation 2.0 ... 7
2.1.3 Storytelling Strategies and Anti-Something Narratives of Extremist Movements ... 8
2.1.4 From Collective to Connective Action ... 9
2.2 Extremist Movements and Social Media ... 12
2.3 Research Questions ... 16
III.
Methodological Approach
... 16
3.1 Object of Study: Pegida ... 17
3.2 Dataset and Research Tools ... 19
3.2.1 Netvizz Search Module ... 19
3.2.2 Netvizz Page Data Module ... 20
3.2.3 Network Analysis ... 21
3.2.4 Image Analysis ... 22
IV.
Results
... 23
4.1 The Movement Pegida ... 23
4.2 The main PEGIDA Page ... 26
4.3 Network Analysis ... 30 4.4 Image Analysis ... 33
V.
Discussion
... 35
VI.
Conclusion
... 44
VII.
References
... 47
VIII.
Appendix
... 53
List of Figures:
Figure 1: Elements of connective and collective action networks ... 11
Figure 2: Page Like Network of Pegida - Crawl Depth 1 ... 30
Figure 3: Cluster one of the Pegida Network ... 31
Figure 4: Cluster two of the Pegida Network ... 32
Figure 5: Cluster three of the Pegida Network ... 32
List of Tables:
Table 1: Top ten Pegida pages on Facebook worldwide by most likes ... 24
Table 2: Top ten Pegida pages on Facebook worldwide ... 24
Table 3: Top 20 posts with the most engagement ... 28
List of Graphs:
Graph 1: Number of posts per month on PEGIDA page ... 27
Graph 2: Engagement per post type ... 27
Graph 3: Development of meme-like content over time (posted by Pegida) ... 34
Graph 4: Growth of engagement per post type in one year ... 34
Graph 5: Content of images on PEGIDA page ... 35
List of Images:
Image 1: Symbol and lettering of Pegida ... 18
Image 2: Heat map of Pegida pages in Europe (by likes) ... 25
Image 3: Heat map of Pegida pages in Europe (by density) ... 26
Image 4: Most popular post on the PEGIDA Facebook page ... 29
Image 5: Tagcloud of most used Hashtags on the Pegida page ... 29
Introduction
I. Introduction
In 1996, the scholars Capitanchik and Whine published a paper about racism on the Internet, stating: “The Internet has revolutionized communications, bringing enormous benefits. But we cannot afford to ignore its negative aspects – the potential to spread ‘hate’ material that not only offends, but seriously threatens racial harmony and public order” (Capitanchik & Whine 1996). The Internet’s characteristics like anonymity, low cost of publishing and its ability to span geographical boundaries make it an ideal tool for extremists to promote their racial views and spread hate speech. Today, almost 20 years later, the discussion about online hate and racism is more relevant than ever, regarding the rise of new far-right populist parties and extreme right-wing movements1 in Europe (Rogers, 2015, 1).
A good example of an extreme right-wing movement rising through the help of the Internet and even more important through social media is Pegida (Patriotic Europeans against the Islamisation of the West). Pegida’s main political ideas and demands circle around the German refugee policy that is seen as a betrayal of the people accompanied by a very strong anti-Islam position. The followers of the original movement from Dresden meet every Monday on the main square in order to protest against the government and the imminent Islamisation of Germany. The movement was founded around one and a half years ago, in December 2014, attracted more than 100.000 follower on the original Facebook page PEGIDA2 within the first
couple of weeks and spreaded its offshoots over the borders of Europe to the United States and Canada (Munzinger 2016a). Questions arising from this phenomenon are: How can a movement grow so rapidly on Facebook and especially spread so wide? To what other political and sociological actors is Pegida connected online and do they build a network of hate? What storytelling and narrative elements does PEGIDA use in order to mobilise, recruit and radicalize followers? And can this narrative be described as a new populist storytelling concept that differs from traditional far-right or neo-Nazi notions?
Although various studies (Bennett & Segerberg 2012; Gerstenfeld et al. 2003; Von Behr et al. 2013) have investigated how extremist groups increase their recruitment, mobilisation and
1 The shifts in European politics are often grouped together under the label of “populist extremist parties”,
“the new far-right”, or “extremist right-wing movements” (Bartlett et al. 2011, 25). Those can be defined as groups with “a deep concern about maintaining national (and sometimes European) identity, which they see as under threat from high levels of immigration—especially from Muslim majority countries unmitigated failure and argue for restrictions on new immigrants, greater emphasis on integration and assimilation and, in some cases, repatriation of immigrants who ‘refuse to integrate’ or whose cultures are seen as in conflict with Western European values“ (Bartlett et al. 2011, 25).
2 For easier distinction between the general Pegida movement and the most popular Pegida page on
Facebook I will continue to spell the Facebook Page from Dresden PEGIDA in capital letters and the general movement Pegida in lower case letters.
Introduction
radicalisation activities online, what narratives they use and to what extent a network between those groups exists, less attention has been paid to the strategies of right-wing extremist groups on social media and only very little research has been conducted so far about the dynamics of Pegida, since the movement has only recently appeared on Facebook. Studying Facebook in order to gain insights that help to explain how extremist movements grow is important since fighting hate speech and extremist groups online is only possible through understanding their dynamics and strategies. Furthermore social media is not only used for the purposes of radicalisation, mobilisation and recruitment but generally transforms the structure of protest into what Bennett and Segerberg call “connective action movements” (Bennett & Segerberg 2012). Those new forms of protest are not just structurally changed through social media platforms but also characterised by attracting people that are not necessarily very political but see protests as an expression of personal hopes, lifestyles and grievances (Bennett & Segerberg 2012, 743). The notion of connective action was so far only applied to slacktivism or leftist protests like the G20 demonstrations or the occupy movement (Dijk & Poll 2013; Segerberg & Bennett 2011) while studies that investigate if there is something like a right-wing logic of connective action are lacking. With this thesis, I will fill the gap in the literature about the fast growing movement Pegida by investigating whether the notion of connective action can also be applied to right-wing movements using Pegida as example and by evaluating how other factors like the echo chamber linking theory influence those logics of action. In order to do so I will use a digital methods approach to examine what studying the web can contribute to better understand Pegida’s dynamics. To find out what narratives Pegida uses to recruit, mobilise and radicalise followers and detect a network of the right on Facebook, the Digital Methods Initiative Tool “Netvizz” will be used for creating a page like network around the main PEGIDA page to investigate whether there is an echo chamber like structure. Furthermore an image analysis of the content posted on the PEGIDA page will be conducted to indicate and evaluate the role of meme-like pictures that are easy to personalize in fast-growing online movements.
Following this introductory chapter, the next part of the thesis will give an overview of the existing literature by discussing, linking and critically evaluating the notion of the echo chamber linking theory, the characteristics of recruitment, mobilisation and radicalisation strategies of extremist movements, the storytelling and narratives of populist right-wing movements online as well as the concepts of collective and connective action. In the second part of the literature review, the focus is set on the existing studies that examine to what extent those theories apply not just online but especially within social media networks to develop a good understanding of the platform affordances of social media platforms and the way, extremist movements use and profit from them. In part III the methodological approach is presented, beginning with a detailed overview of the object of study, the German anti-Islam movement Pegida. The research that was conducted follows a social network analysis approach with an emphasis on a quanti-qualitative
Literature Review
content analysis of the most popular posts as well as an image analysis in order to gain insight into PEGIDA’s media strategy. In chapter IV the results of the research are presented before discussing Pegida’s narrative strategy, the use of Facebook for recruitment, mobilisation and radicalisation and to what extent a right-wing logic of connective action can be determined, in part V of the paper. The findings will be evaluated, interpreted and set into contrast where appropriate with the studies introduced in the literature review. Section VI submarines the results of this work as well as draws and reflects on the conclusions that were outlined before.
II. Literature Review
2.1
Notions of Extremist Movements Online
The beginning of cyberhate3 is generally related back to 1995, more precise to the launch of the
web page Stormfront by the white supremacist Don Black4 (Perry 2000, 121). Stormfront was
not only one of the first hate pages that preached white supremacist ideologies but also the most central site in terms of incoming links and possessed a particular high betweenness-score, meaning that the incoming and outgoing links are widely distributed across the network (Burris et al. 2000, 223). Shortly after the first hate pages went online researchers started concentrating on the topic. Hence the research on cyberhate and extremist movements online hast its beginnings in the mid-1990s and shows a variety of approaches. One of the first examples of studying hate online is an article in the journal of the Institute for Jewish Policy Research by Michael Whine and David Capitanchik in 1996. The authors claim that the Internet – apart from the enormous benefits it brings – also bears the potential to spread hate material that not only offends but seriously threatens racial harmony and public order because it provides far-right groups with the means to communicate, organize and distribute their ideology (Capitanchik & Whine 1996).
According to Hale (2012), the Internet is a popular tool for extremist and hate-groups for several reasons. Mainly because of “easy access, limited or no regulations, huge audiences, anonymity of communication with like-minded individuals, a multi-media environment, and the ability to shape traditional mass media coverage“ (Hale 2012, 344). A report from 2011 of the Simon Wiesenthal Center in Vienna suggested that there are about 14.000 social networking sites, forums, twitter, blogs and newsgroups supporting hate-motivated extremist groups (Hale 2012, 345). The Internet has facilitated the spread of hate messages that typically involve “harassment, bullying, discrimination or other degrading treatment of people” (Perry & Olsson
3 Definition of “cyberhate“: Hatred expressed by means of computer networks (Yourdictionary) Cyberhate
was a common term in the 90s and is now not so popular anymore and often replaced by the term “hate online” (Coe 2015).
4Don Black is a former Klansman from Alabama who succeeded David Duke as head of the Knights of the
Literature Review
2009, 195) who usually belong to a group of a certain race, colour, ethnicity, gender, sexual orientation, nationality, religion or other characteristics (Warner & Hirschberg 2012, 19). In addition, the combination of being anonymous online and the lack of legal regulations make the expression of harassment or hate against minorities or individuals easier (Perry & Olsson 2009, 195). In other words, the Internet creates virtual spaces where homogeneous sentiments of racism and hate can be freely shared without consequences (Perry & Olsson 2009, 191). However Perry and Olsson emphasize, “we have to remember that technological systems are socially produced and that social production is culturally informed and that the Internet is not an exception in this context” (Perry & Olsson 2009, 188). Put differently, Perry and Olsson argue that people are not determined by technology but that the Internet is only a social production that is influenced by the respective culture. Consequently, the phenomenon of hate groups online is therefore only a reflection of hate groups offline. The Internet might function thereby as facilitator or amplifier but is not cause for the emergence of extremist movements.
2.1.1 Echo Chamber Linking and the Creation of a Network of Hate
As a consequence of the new possibilities enriched by the Internet like communication and distribution of hate material other scholars emphasized the risk of the emergence of a “global racist subculture” through the creation of an international collective identity (Perry 2000, 123). In the year 2000 Burris et al. found evidence for the contribution “in the creation of an international virtual extremist community” (Gerstenfeld et al. 2003, 30) with over two third of links between supremacist websites lead to international sites. Several other scholars found evidence that extremist sites link to each other (Burris et al. 2000; Gerstenfeld et al. 2003; Zhou et al. 2005) and showed therefore that the phenomenon of linking was an important factor within the rise of extremism online. According to Perry (2000), one of the main consequences of the linking between those movements or groups is the creation of a “collective identity” (Perry 2000, 123), the common consciousness of belonging to the same group. Especially smaller or geographically isolated groups have the possibility to be linked with bigger groups and become therefore part of the collective. This not only leads to better communication but also strengthens the sense of community (Gerstenfeld et al. 2003, 40). Even though some of those groups or movements might be very different, they all share the same view or beliefs expressed by images and texts ranging from anti-Jewish, anti-immigrant, anti-minority, anti-liberal, anti-gay and so on. The united presentation of those subgroups makes them appear more powerful and less extreme at the same time (Gerstenfeld et al. 2003, 40). According to the scholars this leads to a trivialisation of extreme ideologies: “It can help convince even the most ardent extremist that he is not alone, that his views are not, in fact, extreme at all“ (Gerstenfeld et al. 2003, 40).
Literature Review
What makes them appear less extreme to themselves actually has the opposite effect because they are caught in a so-called “echo chamber” (Key 1966). This notion, first introduced by Key in 1966 describes the amplifying effects of opinion forming. The term serves nowadays as a general metaphor for the effect that the Internet tends to encourage individuals to sort themselves into groups of like-minded people (Jenkins 2008, 195). This theory is based on Sunstein’s notion of “group polarization” (Sunstein 2002, 176), that implies that members of a group with extremist tendencies move toward a more extreme point because, firstly, they keep hearing single-sided ideas and secondly, arguments and group members tend to adopt their peers’ positions and higher collectivity leads to greater confidence, which is in turn linked to more extreme positions (O’Hara & Stevens 2015, 404). This not only applies for groups consisting of members with extremist tendencies but also for groups that share a collective identity like the same political orientation for instance (Sunstein 2002, 176). As a consequence, people will not be confronted with contrary ideas and shift to more extreme points. Sunstein emphasizes that this is not necessarily a bad thing but can lead to dangerous extremism and intolerance against contrasting views and finally limits deliberative, democratic conversation by encouraging ‘‘enclave extremism’’ (Sunstein 2002, 186).
The echo chamber linking theory takes on this notion and is vastly visible in social media, when one user creates a post and his friends or followers take on the opinion and subsequently post something similar. What needs to be stressed here is that social media networks can be huge and therefore many connections echo information and opinions. As a consequence the impression might be created that the vast majority of people already have the same opinion. The opinion adoption process is changed because “cognitive processes can be replaced by network effects in social media opinion spreading“ (Pfeffer et al. 2013, 123). Meaning that people will not actively decide for opinion A or B but are influenced by network effects that make a good – in this case particular opinion – more valuable because many people spread it. Since hardly opposite opinions pass through the walls of the echo chambers, the chamber strengthens and reinforces existing beliefs. “People’s views become more narrow and extreme when echo chambers buttress them“ (Weinberger 2011, 4). Furthermore these reinforcing ideas can create an environment where extreme opinions that are not socially acceptable become the norm (Hirvonen 2013, 81). Supported by omnipresent technologies that filter and recommend mainly those things that people were already interested in before. This creation of a unique universe of information is known as the so-called “filter bubble” (Pariser 2011). The concepts of the filter bubble and the echo chamber are insofar intertwined, as they both influence the decision making of individuals by filtering out unwanted or opposing opinions. According to Pfeffer et al., the filter bubble dominates the information flow by filtering out things whereas the echo chamber additionally guarantees efficient persuasion and affirmation (Pfeffer et al. 2013, 123).
Literature Review
However both concepts contribute to the reinforcement of existing beliefs and opinions by repelling opposing ideas.
Previous studies have proven that users tend to read and link to like-minded resources, like Hargittai et al. who researched the blogosphere around conservative and liberal bloggers and came to the conclusion that political bloggers are much more likely to link to others who share their political views (Hargittai et al. 2008, 85). When it comes to hate groups, researchers proved that they are especially intolerant to other opinions and conspiracy theories are sheltered from objective criticism, while evidence is presented selectively (Kay 2011; Sunstein 2006). Furthermore several scholars (O’Hara & Stevens 2015; Von Behr et al. 2013) proved that the echo chamber thesis could be especially applied in the context of online radicalization, mostly applied to Islamic extremism but similarly connected to all kinds of extremism, such as neo-Nazism (O’Hara & Stevens 2015, 402).
Nevertheless there are several scholars who dissent with Sunstein’s notion of “group polarization” and the echo chamber theory and concluded studies that show opposite developments (Bakshy 2012; Hargittai et al. 2008; Jenkins 2001). One of the counterarguments implies that individuals are usually not just member of one group but take part in several different collectives (Jenkins 2001). Peter Jenkins, who submitted a review of Sunstein’s “Republic.com 2.0” criticized that the trend towards blog aggregation services, which provide a convenient review of the range of opinions on various blogs is completely left out in the group polarization theory (Jenkins 2008, 198). Hargittai et al. conducted a study in 2008 where they found out that there is much more linking between bloggers who share the same ideas. However, at the same time there is also a considerable amount of linking to bloggers with opposing point-of-views (Hargittai et al. 2008, 85). Another study that generated interesting insights in terms of echo chamber building was published in 2012 from the researchers of Facebook’s data team. They found:
That even though people are more likely to consume and share information that comes from close contacts that they interact with frequently [...] the vast majority of information comes from contacts that they interact with infrequently. These distant contacts are also more likely to share novel information, demonstrating that social networks can act as a powerful medium for sharing new ideas, highlighting new products and discussing current events“ (Bakshy 2012).
In other words: Although we are more likely to share information of our close friends, we receive new information from our “weak ties” and they provide access to information one would not have seen otherwise and are therefore indispensable to our network (Manjoo 2012). This can be seen as a proof that people do not get caught easily in groups with similar opinions but that information is received from all kinds of other directions and therefore weakens the echo chamber argument.
Literature Review
2.1.2 Recruitment, Mobilisation and Radicalisation 2.0
Another feature of the Internet that needs to be considered while studying online extremist movements is its effectiveness as a recruiting tool. Traditionally, the primary means of recruiting members was by word of mouth or by distributing printed material (Perry & Olsson 2009, 189). Gerstenfeld et al. found that extremist groups rely heavily on the Internet as a recruiting tool and that special attempts are made to attract a younger audience by using multimedia content such as music (Gerstenfeld et al. 2003, 39). The main function of the Internet is the communication factor, since it offers various ways for people, especially isolated individuals, to get in contact with certain groups (Gerstenfeld et al. 2003, 39). Within the supremacist movement soft-core sites might serve the more extreme groups in their recruitment since they could function as important ports of entry into more radical movements (Burris et al. 2000, 232).
Furthermore the mobilisation aspect needs to be evaluated in more detail. The so-called resource mobilisation theory emerged within the study of social movements and argues that voluntary organizations need to mobilise resources to attract and recruit participants. Resources can be knowledge, money, hours of labour but also solidarity or legitimacy (Duijvelaar 1996). According to Hara and Estrada computer-mediated communication can be also included as a tool that can mobilise a variety of participatory resources, especially “knowledge, interpersonal interactions, identity support, and the building of credibility and legitimacy. “ (Hara & Estrada 2004, 505) Knowledge is always accessible regardless of the location or time zone of the user and builds therefore a basis for further mobilisation. Knowledge access is combined with the increase of interpersonal interactions with like-minded people that leads to identity support and the validation of personal views (Hara & Estrada 2005, 504). Those elements of mobilisation can be referred back to the notion of echo chamber linking where opinions are reinforced through the lack of opposing views because like-minded individuals or actors will echo the already existing beliefs. Hate groups furthermore use the Internet to establish and spread legitimacy and at the same time increase visibility and credibility whereas the attention of watchdog groups can be seen as successful legitimation and the establishment of credibility because they get treated serious by other actors (Hara & Estrada 2005, 504).
In addition to the Internet’s influence on recruitment and mobilisation activities comes the radicalization aspect. Radicalization describes the process whereby groups or individuals become political extremists (Neumann 2013, 454). Von Behr et al. identified five hypothesises how the Internet influences the radicalisation process. Firstly the study supports the assumption that the internet creates more opportunities to become radicalised because the Internet functions as key source of information and communication combined with the connection to like-minded individuals that are all victims of propaganda for their extremist beliefs (Von Behr
Literature Review
et al. 2013, 17). Secondly, the Internet connects not only to like-minded people or actors but also creates a good opportunity to confirm and reinforce existing beliefs. Already extremist beliefs could therefore become even more extreme as individuals find their views echoed by other like-minded individuals and the Internet functions as some sort of echo chamber (Von Behr et al. 2013, 19). The Internet has furthermore being referred to accelerate the radicalisation process by several scholars whereas the study conducted by Von Behr et al. came to the conclusion that the Internet does not necessarily support this hypothesis. But what is for certain is that the Internet enables or facilitates the radicalisation process (Von Behr et al. 2013, 19). Other interesting findings generated by the study were that physical contact is not replaceable by online interactions but rather complements communication aspects (Von Behr et al. 2013, 20). As well as that the contact to other individuals still plays an important role within radicalisation processes and that the Internet does not increase opportunities of self-radicalisation (Von Behr et al. 2013, 20). Online radicalization, or cyber extremism is a growing concern to the society and also of great pertinence to governments and law enforcement agencies (Correa & Sureka 2013, 3) and especially the use of social media tools by individuals and organizations to radicalize followers for political and social change has become increasingly popular (Thompson 2011, 167). The notion of extremism on social media platforms will be further discussed in the last part of the literature review.
2.1.3 Storytelling Strategies and Anti-Something Narratives of Extremist Movements
The use of narratives to influence individuals and promote ideas plays a major role within hate groups. The use of narratives and storytelling strategies can help to elicit fewer counterarguments and less resistance to persuasion (Lee & Leets 2002, 928). The main strategy of all kind of movements or political parties is the so-called persuasive storytelling, which means the use of a certain narrative to persuade or convince instead of having simply entertainment or enlightment as communication goal (Lee & Leets 2002, 929).When it comes to nationalist populist movements and parties in particular, Jansen describes them as “large scale political project that mobilizes ordinarily marginalized social sectors into publicly visible and contentious political action, while articulating an anti-elite, nationalist rhetoric that valorises ordinary people” (Jansen 2011, 82). A strong emphasis lies thereby on “the people“ which mainly implies ordinary members of the national community (Jansen 2011, 84). Those people are placed in contrast to the anti-popular elite that can be identified as economic or political ruling powers that are portrayed as having unjustified control over the ordinary citizens (Jansen 2011, 84). A third aspect is the targeting of certain anxieties that exist within society caused by complex phenomena like globalization or integration with a strong “anti-something” narrative (Mergenthaler 2015, 7). For instance populist and far-right
Literature Review
groups provide a narrative responding to fears related to identity and affiliation by using a strong ant-immigration or anti-minority rhetoric (Mergenthaler 2015, 12). This can be lead back to a general politic of blame instead of looking for political solutions that those parties or movements pursuit (Mergenthaler 2015, 8). Those strategies taken together form a populist rhetoric that implies symbolic actions, styles of expression and public statements and not just justifies but also animates political action in the real world (Jansen 2011, 83). Narratives and storytelling elements are therefore a fundamental part of mobilising and activating people to come together and demonstrate for their ideas or demands and therefore also count as important factor for logics of actions that will be further described in the next section.
2.1.4 From Collective to Connective Action
Bringing people together in order to achieve some kind of public good, e.g. a democratic reform is the main subject of social movement theory. Olson, who researched the dynamic of groups, introduced the term “collective action“ (Olson 1965) that underlies the assumption that especially large groups have problems to efficiently act together because rational individuals will free-ride on the efforts of others within the group (Bennett & Segerberg 2012, 749). Furthermore social mobilization depends on collective identification and action frames that can only be assured through high hierarchical structures (Dijk & Poell 2013, 8). The organizations involved play an important role bridging differences in opinions and ideas and the emphasis is on the creation of interpersonal networks in order to build relationships within the participants of a movement. The main elements of collective action movements are therefore the creation of interpersonal networks in order to strengthen the collective identification, high hierarchical structures, a strong organizational coordination and the emphasis on getting the people to contribute to the collective (see Figure 1).
Bennett and Segerberg spin the notion of collective action further and argue that a new logic of self-organizing movements, developed as a consequence of communication technology, emerged alongside collective action networks and shift into a phenomenon what the scholars call “connective action” (Wright 2015, 416). At the core of the logic of connective action is the fact that digital media function as “organizing agents”, what forms the first important characteristic (Bennett & Segerberg 2012, 752). That means that connective action networks organize themselves through social media, which leads to the consequence that strong hierarchies, and collective identities are no longer needed (King 2014, 967). The second main element digitally mediated networks is the notion of self-motivated participation as personally expressive content is shared with friends and family in the network. Within connective action movements, self-motivated contribution to a common good becomes a way of personal expression, which is the third main element of connective action movements (Bennett &
Literature Review
Segerberg 2012, 752). Personal expression is the starting point and is carried out through the sharing of already personalized ideas, plans images and resources in social networks like Facebook, Twitter and YouTube (Bennett & Segerberg 2012, 753). In order to provide easily transferable and customizable memes and messages that can be personalized, it is important that those action frames are open to interpretation, so that many people can support the message for different reasons (Wright 2015, 424). An example for those easy-to-personalize memes and action frames that quickly spread over the social network is the action frame “We are the 99 per cent“ that was used during the occupy protests. This frame got shared via personal stories and travelled through the web e.g. via Tumbler, Facebook and Twitter during the protests (Bennett 2012, 742). People personalized the catchphrase „We are the 99 per cent“ by sharing pictures of them, holding a paper with their handwritten story on it, explaining their struggling with student loans, the housing market crash or the lack of health insurance, stating: “I am the 99 per cent“. This example shows that the meme “We are the 99 per cent” is open to interpretation because people were personalizing and spreading it for all kind of different reasons.
Bennett and Segerberg furthermore distinguish between two forms of connective action: self-organizing networks and organizationally enabled networks (see Figure 1). Within the former there is only little or no organizational coordination involved, the participation of formal organizations is shunned and personal expression is shared over social networks without moderators interfering (Bennett & Segerberg 2012, 742). Bennett and Segerberg give the example of the los Indignados5 protests for a self-organizing network. No major organizations
appeared at the centre of the movement and the little amount of formal actors involved stayed in the background to “honour the personalized identity of the movement” (Bennett & Segerberg 2012, 741) that was represented by the many ordinary people that were affected by financial and political crisis. Plus the easy to adopt personal action frames that travel online and offline through the help of online platforms like the media hub of Democracia real YA!3 that included over 80 local Spanish city nodes and a number of international solidarity networks, make this movement an ideal example for a self-organizing network (Bennett & Segerberg 2012, 755). Whereas in organizationally enabled networks, loose organizational coordination is involved. This involvement leads to some organizational moderation of personal expression in social networks (see Figure 1). Here the occupy6 movement serve as a good example. Even though the
movement spread under the easy-to-personalize frame of “We are the 99 per cent” there was
5 Los indignados (the indignant ones) mobilized protesters in over 60 cities in Spain in the year of 2011
under the name 15M (May 15) when the portests took place. The sustained protest organization managed to keep social parties, unions and other formal organizations out (Bennett & Segerberg 2012, 741).
6 The occupy Wall Street protest took place in the United States in 2011 and spread in a month from New
York to over 80 countries and 900 cities around the world combining the participiation of some formal organizations and loose-tie participants via technology (Bennett & Segerberg 2012, 741).
Literature Review
some organizational development displayed (Bennett & Segerberg 2012, 755). For example the General Assemblies with regular attendance, division of labour and coordination of actions and layers of technologies and websites like the website 15.10.11 united for #global change that asks “who are you?“ instead of “who we are“ (Bennett & Segerberg 2012, 754). Apart from those formal organizations there was a big endeavour not to focus on leaders or official spokespeople but using the technological platforms to personalize the movement wherefore it was still defined by its self-organizing roots.
Figure 1: Elements of connective and collective action networks (Bennett & Segerberg 2012, 756)
To briefly sum up the two concepts of connective action displayed in Figure 1, one can say that self-organizing networks build on personal action frames and function without or only little central or lead organizational actors by using social online networks as important organizational agents instead (Bennett & Segerberg 2012, 755). Whereas organizationally enabled networks can be seen as hybrid pattern between the self-organizing networks and collective action movements that are organizationally brokered. In this middle type, formal organizational actors can be involved but only in the background and are characterized by their loose, more entrepreneurial relation with their followers (Bennett & Segerberg 2012, 757). Those two types of connective action stand in contrast to the notion of collective action whereby strong organizational coordination and organizational management of social networks are combined with an emphasis on collective identity and action frames (see Figure 1).
In order to achieve those logics of action and get people to participate in protests, the notion of echo chamber linking plays an important role. Echo chamber like networks can also be described as hyperlink networks since the different actors link to each other, or in the case of Facebook like each other’s pages. Striking within those networks is the high homophily that is
Literature Review
created through those links or the liking that enable members and non-members to reach “a homogeneous set of like-minded organizations in order to enhance visibility of network’s goals” (Shumate & Lipp 2008, 179). Within echo chamber environments with high homophily, collective identity is reinforced because opposing views are banned (Perry 2000, 123). The lack of diverse opinions combined with the reinforcement of the individual views strengthens the commitment to the ideological group (Boutyline & Willer 2015, 3). Strong commitment within groups leads in return to more social pressure to comply the behavioural norms and therefore increases the social pressure to take part in protests for instance (Boutyline & Willer 2015, 3). To sum up, it can be said that echo chamber linking is likely to increase the participation in political action. In the case of collective action networks especially because high homophily reinforces existing views by excluding differing opinions. This leads to a stronger conviction that those views and ideas are normal and therefore instigates political action more easily. When it comes to organizationally enabled connective action networks, political capacities depend amongst others on the stability of the organizations sharing communication linkages and the volume of personal engagement (Bennett & Segerberg 2012, 761). Since organizationally enabled networks are depended on a loose network of organizations, an echo chamber linking structure is very helpful in order to generate political participation. High homophily and a dense network between actors that work together will help to provide not a platform but a network of platforms to organize the self-motivated participation in protests.
2.2 Extremist Movements and Social Media
As described in the paragraphs above, many scholars have examined the notions of group polarization and echo chamber linking (Gerstenfeld et al. 2003; O’Hara & Stevens 2015; Sunstein 2002), recruitment, mobilisation and radicalisation (Gerstenfeld et al. 2003; Hara & Estrada 2005, Neumann 2013) storytelling and narratives of extremist movements (Lee & Leets 2002; Furlow & Goodall 2011) as well as the concept of connective action (Bennett & Segerberg 2012; Dijk &Poell 2013; Wright 2015) whereas only little is known about how these concepts show off in social media environments and how extremist movements operate and benefit from social media networks like Twitter or Facebook.
Echo Chamber Linking
Several aspects make social media an effective tool for extremist movements to radicalize opinions and attract new followers. According to Thompson, “social media easily connects people with a wide audience, the synergy creates a movement en masse of like-minded persons“ (Thompson 2011, 175). Through the creation of social groups consistent of people that are like-minded, existing views get confirmed and opinions eventually become more extreme (Bartlett
Literature Review
2014, 108). This can be referred back to the echo chamber theory, assuming that linking to like-minded people facilitates radicalization and social media therefore serves as an amplifier of extreme views. The topic of political homophily leads to the well-known conflict that digital media either lead to the formation of a public sphere that facilitates the exchange of different opinions and information or exactly the opposite, digital media building an echo chamber that reinforces already existing perspectives and opinions leading to more extremist and narrow minded views (Colleoni et al. 2014, 317).
Research on Twitter that was performed so far generated contrasting results, showing that social media tends to foster the public sphere scenario with low homophily (Kwak et al. 2010) as well as the echo chamber scenario where homophily is high (Conover et al. 2011; Weng et al. 2010). Next to several other scholars (Conover et al. 2011; Kwak et al. 2010; Weng et al. 2010) Colleoni et al. studied homophily on Twitter as well and came to the conclusion that a distinction must be made. Twitter as a social medium shows higher levels of homophily and a more echo chamber like structure. Whereas Twitter as a news medium has more public sphere-like characteristics with lower levels of homophily (Colleoni 2014, 328).
Recruitment, Mobilisation and Radicalisation
Only a small amount of publications appeared in recent years documenting research about recruitment, mobilisation and radicalisation on social media (Bartlett 2011; Dijk & Poell 2013; Theocharis 2011; Thompson 2011). However, when it comes to recruitment, there is not much evidence on how social media encourages the process. One of the few studies that exist suggests that there are two main processes – the dynamics of information diffusion and the dynamics of recruitment – taking place parallel when it comes to mobilizing a huge number of people through social online networks (González-Bailón et al. 2011, 1). Within the process of recruitment, there is no need of being central in the network since the individuals that spread topologically heterogeneous messages early will create random seeding within the network (González-Bailón et al. 2011, 2). Furthermore horizontal organizations are especially successful at recruiting people through social media because they have a decentralized structure that is:
[…] based on coalitions of smaller organizations, plant activation seeds randomly at the start of the recruitment process, which maximizes the chances of reaching a percolating core; users at this network core, in turn, contribute to the growth of the movement by generating cascades of messages that trigger new activations, and so forth. (González-Bailón et al. 2011, 2)
This process embodies the bridges between global and local networks and is also the reason why Twitter played such an important role in many recent movements and protests since it combines the global reach of broadcasters with local, personalized relations (González-Bailón et al. 2011, 2).
Literature Review
Facebook, Twitter and YouTube are stated to have momentous influence in the process of mobilising followers (Dijk & Poell 2013, 11). Theocharis conducted a study where he explores students’ use of websites and Twitter during the 2010 university occupations in the UK and showed how those protests serve as an example of how digital media facilitate political activism and mobilisation (Theocharis 2011, 165). Those students used social media to promote their aims, coordinate the offline protests and mobilise others (Theocharis 2011, 165). The use of social media implies mainly informing and mobilising others by exchanging strategies or information, spread messages of solidarity, update followers on past, current and future protest events, provide links to information, invite people to the protests, convey the feeling of the occupation but also send tweets that were humorous (Theocharis 2011, 179). Furthermore the significant reduction in costs of mobilizing followers, but also organising offline protest and communication in general are currently helping populist parties enjoy dramatic levels of support and prominence (Bartlett 2014, 101).
Social media also influences the process of radicalisation, whereas Thompson argues for instance that social media is not a cause for radicalisation or revolutions, but it has a strong influence on their effectiveness, especially in countries with a strong Internet presence (Thompson 2011, 179). In order to radicalise individuals a combination of high density of connections, a high density of information sources and a strong emotional connection to the event are important (Thompson 2011, 177). Social media functions as hub by connecting people with different sources of information and brings the event closer to the individual, which can lead to a strong emotional involvement and to more radical actions (Thompson 2011, 177). Due to those affordances the probability of a release of political action is higher and therefore it can be said that social media makes radicalisation more effective and therefore enables, causes and accelerates radicalisation to a certain extent.
Connective Action
Recently a lot of mobilization took place in large numbers around topics like privacy, environment or social issues. The interesting thing about those mobilization campaigns is that they were organised outside of traditional political party activism and followed a new logic of social-media based communities, what can be described as connective action (Mergenthaler 2015, 7). Social or digital media play a crucial role as organizing agents in those modern protests. According to Bennet and Segerberg, digital media do much more than just simplify the communication and support the organization of protest but rather become actors themselves through shaping the structure of such mobilizations (Mattoni 2015, 1). Important hereby is to emphasize that social media alone is not what makes the difference between collective action and connective action networks since digital media is used in both concepts in order to facilitate communication. The difference is that digital media change the action in connective action
Literature Review
networks structurally since they enable participation as self-motivated and make personal expression of ideas and opinions possible through sharing, commenting and liking (Bennet & Segerberg 2012, 753).
The personalization aspect contains of two main elements. Firstly political content in form of easy-to-personalize frames like the “We are the 99 per cent” example that help to bridge differences of how people feel about certain ideas and political content because it is free to broad interpretations and unites individuals with different personal reasons for wanting to change a situation (Bennett & Segerberg 2012, 744). Secondly the communication technologies that enable the spread of those personalized memes via social media networks (Bennett & Segerberg 2012, 745). When Bennett and Segerberg refer to an easy-to-personalize action frame or meme they mean “a symbolic packet that travels easily across large and diverse populations because it is easy to imitate, adapt personally and broadly with others. [...] They travel through personal appropriation, and then by imitation and personalized expression via social sharing in ways that help other appropriate, imitate and share in turn“ (Bennett & Segerberg 2012, 745). Those action frames need to be easy to shape and share so they can travel interpersonally through newspapers, blogs and Facebook friend networks. Collective action frames that are not so easy to adapt for broader interpretation and are more depending on sharing common identifications can also be regarded as memes if they imply some sort of social transmission like the slogan “eat the rich“ (Bennett & Segerberg 2012, 747) that refers to Jean-Jacques Rousseau’s statement during the French Revolution: When the people shall have nothing more to eat, they will eat the rich.
Concrete examples for those new forms of organisation and protest are the 2011 occupy movements where technology-enabled personal networks were not only used for communication purposes but also “empowered flexible organizations that allowed rapid action and coordinated adjustments“ (Dijk & Poell 2013, 8) and the les indignados movement in Spain where the mobilization effort was supported by the use of social networks that have produced a personalized digitally networked action (Anduiza et al. 2014). The whole les indignados movement took place outside the reach of political parties and unions and the participants had low levels of previous political activity (Anduiza et al. 2014). The movement became so successful because many small organizations networked through social media in order to achieve a common goal (Mergenthaler 2015, 7). Another example is Segerberg and Bennett’s study where the scholars analysed two Hashtags used in the protests around the United Nations Climate Summit in Copenhagen in 2009 and approved that Twitter functioned as an organizing mechanism or organizational agent in these protests (Segerberg & Bennett 2011, 212). Furthermore they detected different roles of the Hashtags within the protests. The dynamics of #thewave can be described as a driver for mobilization and publicity resources set by some central actors, whereas #cop15 was showed more self-organized stream characteristics and
Methodological Approach
therefore seek resources in different ways from organization-centred streams and fit therefore in the notion of connective action where no hierarchical structures are needed (Segerberg & Bennett 2011, 213).
2.3 Research Questions
Further research in the field of how extremist movements use social media is desirable, since the increasing presence and enormous power of social media has appealed to extremist organizations and individuals to attract new followers (Correa & Eureka 2013, 3) For that reason I will set the focus of my research on how extremist movements unfold their hate and propaganda messages on Facebook, setting the focus especially on the rapid growth of the anti-Islam movement Pegida. My research questions therefore read as the following:
(1) To what extent can an echo chamber linking structure be detected between the respective Pegida pages and the other actors connected to the movement on Facebook?
(2) What kind of narrative and media strategy does PEGIDA use in order to recruit, mobilize and radicalize followers on Facebook?
(3) When looking at PEGIDA, to what extent can be said that there is a right-wing logic of connective action?
III. Methodological Approach
The methodological approach that was used in order to generate the results of this thesis will be described in the following section. Instead of using the traditional empirical methods like surveys, interviews and experiments this study follows a digital methods approach, meaning techniques for the study of societal change and cultural condition with online data (Rogers 2015, 1). When it comes to the study of the web there was a shift within the last years from seeing the Internet as a separated virtual society to an approach where Internet-based data is seen as a means that could revolutionize the understanding of society at large (Rogers 2015, 2). Digital methods make thereby use of the affordances and functions that are built into the devices such as Google Web Search and Facebook’s Graph Search like hyperlinks, tags, shares and rewets (Rogers 2015, 1). One of the characteristics of conducting research with digital methods is that there is no active participation of the research subjects needed. However, digital methods also face different problems because they developed in tandem with the medium conditions and are built on top of other devices (Rogers 2015, 8) like the Netvizz Search Module that is an interface
Methodological Approach
to Facebook’s standard search function and therefore sits on top of the Facebook API. Problems appearing related to web data can vary from the impermanence of web services and the consequential instability of data streams, as certain services are discontinued, as well as the unreliability of the quality of web data that can be messy, unstructured or incomplete (Rogers 2015, 2). Furthermore, there is always a distinction between researching the medium and societal research. Medium research means thereby the study of how and for whom a platform works whereas the information retrieved from the digital methods approach is used as indicator of social trends. Especially social networking sites provide rich opportunities for social and cultural research, providing different sorts of insights into their composition and characteristics (Rogers 2013, 35).
In order to gain a deeper understanding of the dynamics of Pegida on Facebook I will use a combined approach since I want to study the social phenomena of the characteristics and success of Pegida as well as look into how the affordances of social media platforms like Facebook change the way of protest as well as help a movement to spread and grow. The following section will contain a detailed step-by-step description of the research that was conducted, starting from gathering the data via the Netvizz application to the performance of a page like network analysis and an image analysis using Microsoft Excel, espatial, Gephi as well as Adobe Photoshop to visualize the data.
3.1 Object of Study: Pegida
The object of study of the following research is Pegida, a German anti-Islam movement that gained huge popularity within the last year.7 Pegida stands for “Patriotic Europeans against the
Islamisation of the West“ and is a right-wing populist movement that combines the fear of Islamisation with general criticism of Germany’s politics and mainstream media (Dostal 2015, 523). The movement was founded in late October 2014 in Dresden where protesters met for the first time for their regular “Montag Abendspaziergäng“ (Monday evening walk). The remarkable thing about PEGIDA is the mobilization of around a dozen citizens demonstrating on the streets that were barely involved with political parties before (Dostal 2015, 523). PEGIDA embodies a right-wing ideology, while insisting on representing the ordinary people, defending the mainstream of German culture and traditions combined with a strong anti-establishment focus (Dostal 2015, 524). In December 2014, a list of nineteen demands was put forward including, among others, the defence of the Christian-Jewish culture of the Occident, a controlled migration policy, a much quicker system of decision-making on political asylum and the expulsion of all
7 When the movement started in October 2015 around 350 people protested in Dresden against the
“Islamisation of the West” whereas at the end of the year around 20.000 people joined the movement according to the police estimates. Even though the numbers vary according to different sources, an explosively increase of the popularity of the movement cannot be denied (Reuband 2015, 133).
Methodological Approach
rejected asylum seekers from Germany (Dostal 2015, 525). Within the claims it was striking that no direct criticism of Islam was expressed in order to avoid accusations for being racist or extreme and to appeal mainstream conservatives (Dostal 2015, 525). In contrast the tone of the Monday rallies is much more radical, supported by guest speakers like the Dutch far-right populist politician Geert Wilders8. Furthermore there is a strong scepticism against the
Lügenpresse (lying press) that is why Pegida supporters were advised not to talk to mainstream media during protests (Dostal 2015, 525). A part from the concerns mentioned above that Pegida followers share the increasing conflict about the refugee crisis in Germany needs to be mentioned here (Dostal 2015, 526). The movement uses a pictogram as its icon (see Image 1) of a person dropping a Swastika, the symbol of „Antifascist Action“, the symbol of ISIS and the Kurdish PKK flag into a rubbish bin (Becher et al. 2015, 8).
Image 1: Symbol and lettering of Pegida
When it comes to the profile of Pegida’s followers it is interesting to observe that they derive from the centre of the society. Seymour Martin Lipset warned of potential “middle-class extremism“ (Lipset 1960) instancing the example of many middle-class Germans’ support for the Nazi movement. He furthermore refers to Marxist notion of the pretty bourgeoisie might turn to the far-right during periods of economic crisis (Dostal 2015, 526). Lutz Bachmann, co-founder and chairman of Pegida started the PEGIDA Facebook page on December 11th in 2015 that was ever since the major mobilizing device of PEGIDA (Dostal 2015, 524). Apart from its success in the region around Dresden and some other places in Germany the movement did develop around whole Europe and even spread in Canada, South America and the U.S. Although Pegida is less successful in most other countries, it is still interesting to see how far and wide the movement spreaded.
In order to understand the dynamics of Pegida and its followers it needs to be researched where it begun – on Facebook. Without the social network platform, the rapid grow of Pegida would be not have been possible, since it is the movement’s most important mobilization platform where they received more than 150.000 likes within the first month (Menninger 2016).
8 The Dutch far-right populist politician Geert Wilders held a speech in front of thousands of Pegida
followers to counter society’s Islamisation but failed to attract as many listeners as the organisations had hoped for (The Telegraph 2016).
Methodological Approach
3.2 Dataset and Research Tools
In the following section the methodological approach that was used to generate the results of the empirical research will be introduced in detail. In the first paragraph the tools that were used will be described, mainly the Netvizz Search Module as well as the Netvizz Page Data Module.9 Furthermore the dataset will be presented. In the second part the methodological
performance of a network analysis and image analysis will be outlined.
As a very first step a new Facebook account for the purpose of the research was created. With the help of the research profile, the following tools were used to conduct the research. The main software that was used is Netvizz, a tool developed by Bernhard Rieder, researcher and developer at the Digital Methods Initiative. Within the Netvizz tool there are several sub-tools. The Netvizz Search Module, the Netvizz Page and Group Data Module, the Page-Like-Network tool and the Link-Stats tool (Netvizz 2012). The Netvizz Search Module can be used in order to understand how a particular topic, in this case the right populist movement Pegida, is covered on Facebook. The Netvizz Search Module is an interface to Facebook’s standard search function and sends the query to Facebook’s normal search feature and eventually obtains a list with basic statistics about the displayed pages (Rieder 2015). The Netvizz Page Data Module gets either the last respective amount of posts or all posts within a particular date range and creates networks and tabular files for user activity around posts on pages. Furthermore the Page Like Network tool can be used in order to create a network of pages connected through the likes between them. In addition to that the visualization tool Gephi10 was used. Gephi helps to find patterns and
clusters within the data and makes network connections visible. With Gephi, data analysts can interact with the representation, manipulate the structures, shapes and colours in order to make hypothesises and discover patterns.
3.2.1 Netvizz Search Module
In order to get a first overview not only of all existing Pegida pages on Facebook, but also of the anti-Pegida movements, it was necessary to create a list of those Facebook pages as a base and starting point for the further research that will consist of an analysis of the general statistics about all Pegida pages and anti-Pegida pages, an examination of the metadata of the most popular Pegida page from Germany as well as a network analysis and image analysis. A list with 511 results was created with the Netvizz Search Module after querying “Pegida”. The following information was provided in the list: (1) Name of the Page (2) Link (3) Hyperlink - I added this column for easier performance (4) Category (Aerospace / Defence, App Page, Artist, Author,
9 https://apps.facebook.com/netvizz/ 10www.gephi.org
Methodological Approach
Blogger, Book, Cause, Church / Religious Organization, Comedian, Community, Community / Government, Community Organization, Education Website, Education, Entertainer, Interest, Local Business, Media / News / Publishing, News Personality, Non-Governmental Organization (NGO), Non-Profit Organization, Organization, Political Organization, Political Party, Public Figure, Regional Website, Society / Culture, Teacher, TV Network, Video Game, Website, Website, Work Position) (5) Likes (6) Check-in (7) Talking about count (8) Website (9) Cover (Picture) and (10) Description. After cleaning the data, including the correction of spelling mistakes, occurred through the HTML mode that changes special characters like German umlauts. Furthermore non-topic related pages were detected and deleted. In a second step the list was divided into pages that are in favour of Pegida and their anti-Pegida counterparts. The finalised list included then 345 Pegida pages and 152 anti-Pegida pages. Unfortunately there is no geolocation given within the data. Therefore the country, region and city of each Pegida and anti-Pegida page were added manually. With this information several lists could be easily created and organised according to various criteria, for instance the most popular pages in Germany, Europe or the world.
3.2.2 Netvizz Page Data Module
In a second step the main PEGIDA page was scraped in order to get more detailed information about what content got shared, what posts are most popular and how the page evolved over time in order to get insights into their media strategy. The most popular page is PEGIDA, listed within the category “Community / Organisation” and with currently (day of research 07.03.2016) 204.781 likes11 (see Appendix 4.1). In the imprint the “PEGIDA Förderverein e.V.”
(PEGIDA booster club) is named as responsible organization with the main address in Dresden, where the Pegida movement has its roots in. Another thing to mention is that only the page administrators are allowed to post on the page. All other Facebook users can only like and comment on the post or share it.
In order to scrape the data of the PEGIDA page the Netvizz Page Data Module was used. The date range was set from 29th December 2014, the day when the first post was made until the
25th February 2016, so all in all 14 months of Facebook activity on the page were scraped. The
module creates four different files including a tabular file that lists different metrics for each post, a tabular file that lists basic stats per day for the period covered by the selected posts, a tabular file that contains the text of anonymized user comments and a bipartite graph file in gdf
11 For a better assessment of the dimensions of the amount of likes: The biggest anti-Pegida page in
Germany “PEGIDA#watch” has 54.504 likes at the moment of investigation. The CDU, the major centre-rightparty in Germany and currently part of the coalition government hast only 102.467 likes on their official Facebook page, thus half of the main PEGIDA page.
Methodological Approach
format that shows posts, anonymized users and connections between the two whereby a user is connected to a post if she commented or liked it. The information of the first tabular file with metrics for each post was used to create an excel file with 2813 posts containing information about the type of the post (event, photo, video, link, status) by whom the post was done, the content of the post, the picture if any, the link domain, the date of the time the post was published, how many likes, comments, comment replies, shares, comment likes and engagement12 in total the post received, as well as the post ID and the post link. With this
information three graphs were created. The first graph shows how many posts the administrators made per month. With a low period in March and April 2015 but high peak from October 2015 on (see Appendix 4.2). The second graph shows what kind of posts were mostly made, which was by far links and photos (see Appendix 4.3) and a third graph containing information about what post was the most popular and what kind of engagement – likes or comments – it provoked (see Appendix 4.4). Furthermore a Hashtag Tagcloud was created. In order to do so, all the Hashtags on the main PEGIDA page were detected and copied in a list with a total of 2993 Hashtags. The data was cleaned by adjusting similar Hashtags like “Pegida”, “Pegida!” and “Pegida!!” to one category with a total result of 662 different Hashtags. The Hashtags were then visualized in a Tagcloud (see Image 5).
3.2.3 Network Analysis
The data extracted from Netvizz was furthermore analysed from a social network analysis perspective. Within this approach a network of likes around the PEGIDA page on Facebook was explored to get a deeper understanding of the relationship between the PEGIDA Facebook page and the pages or groups that the PEGIDA page liked or was liked by them. The Netvizz Page Like Network tool relies on the fact that Facebook pages can like other pages. The tool crawls the network of pages liking each other in order to get a larger ecology of pages and to see what other page is a page affiliated with (Rieder 2015). The only parameter is the crawl depth, hence the value how far the tool goes into the network. The first network that was created has the PEGIDA page as a starting point and a crawl depth of one. The Netvizz Page Like Network tool starts with the seat page and gets all the pages the PEGIDA page likes and all relationships between those pages. A second network was created, again with the PEGIDA page as starting point but this time with a crawl depth of two, where not only the pages the seat page likes were received, but also the pages that the seat pages likes and the pages they like, so a second step into the network was made (see Appendix 5). The data was then visualized with the software Gephi.
12 Definition of engagement according to Facebook: Engagement rate ist the percentage of people who saw
